Questions
The questions that come up on the phone.
Sixteen of them, grouped, in the order they usually arrive. Each answer is a few sentences and a link to the page that covers it properly. Nothing here is behind a click.
How this page is written
Where a thing is built, the answer says what it does. Where a thing is a commitment we are making rather than software that already runs, the answer says that instead and names which it is. Five answers on this page name something that is not built — the review screen, the rebuild comparison, the domain check, the export, and the fact that none of it is running anywhere yet — and each says so in the sentence rather than in a footnote.
We will not answer a question about results with a number. Not now, and not until there is a school willing to be named beside it.
First
Before you send anything
Can I try it before I pay anything?
Yes, and it is the only sensible way to start. Send one chapter — about fifteen pages — and we convert it at no charge. It comes back within five working days, and your side of it is one email and roughly an hour reading the result. Pick the chapter you know best, the one where a wrong sentence would jump off the screen at you.
What does it cost?
A first conversion is a few thousand euros, and what moves the number is mostly how long your book is. It is not a subscription. There is no price list on this site because we would be inventing the tiers, and you will have the figure in writing before any work starts. The chapter you send to find out whether we are any good costs nothing at all.
How long does it take?
The free chapter is back within five working days. A whole book is measured in weeks rather than days, and most of that time is a person reading the output rather than a machine producing it. You get the number before you commit rather than after. If you need it live next week, we are the wrong supplier and we will say so. How it works has the stages.
What if my book is a scan?
Then it goes through text recognition before anything else happens. That step is built and it is switched off by default, because it bills by the page; it is turned on for a real conversion and capped so a runaway job cannot spend your money. What we will not do is guess: a page we cannot read comes back by number with the reason, rather than filled in with something plausible, because a wrong page number under a citation is worse than no citation. Photocopies of photocopies want re-scanning at 300 DPI.
Do you work in my language?
Your course is in whatever language your book is in. We do not translate, and nothing rewrites your sentences into another language — the rule that keeps your questions word for word applies to the language they were written in. German, Croatian and Serbian are the three the internals were built and tested against, including the case a Serbian publisher will recognise: the same manual issued in Cyrillic and in Latin is treated as one document rather than billed as two. This site and the provider console are in English today. Ask before you assume your alphabet is handled — we would rather tell you no now than halfway through.
Second
What comes back
Is this AI?
No language model is called anywhere in this product. There is no code path that reaches one — the machine reads your file and records which page every piece of text came off, and the writing is people. That is slower, and it is most of what you are paying for. The four origin labels exist for the day any of that changes: if a paragraph is ever machine-written, it says so on the paragraph, where your students can see it, rather than in a settings page nobody opens. What is in it shows the four.
Who reads it before I do?
A person. Extraction and page matching are machine work; then somebody sits down and reads the result before it reaches you. The software knows this is meant to happen — a course version has a state called in review and a job can stop and ask for a human rather than fail — but there is no review screen built yet. Today that review is a person with the output in front of them, and we would rather say so than imply a tool.
Can my instructors edit it afterwards?
Your instructors can read the course and see how students are getting on; they cannot change a word of it, and that is deliberate. Editing belongs to whoever you have made owner, admin or editor — and an editor writes but cannot publish, so making a version live still takes an owner or an admin. When one of them changes a block by hand, it is marked as edited by a person and a later rebuild never overwrites it: it flags it as stale for somebody to look at, which is a different thing.
What happens when the regulations change?
You send the new edition the same way you sent the first one. The intention is that only the parts that actually moved are rebuilt, and the machinery for that is in place: every section carries a key that survives a new foreword pushing all the page numbers down by two. The comparison itself is not built. Today it is us working out what changed, so treat it as an undertaking rather than a feature — the second year sets out what we do.
| Origin labels on a block of text | 4 — from your material, written from it, written without a source, written by a person |
|---|---|
| Ways a question can be recorded | 3 — taken from your material, machine-written, written by a person |
| Question types | 4 — single choice, multiple choice, true or false, free text |
| Kinds of block a lesson is built from | 11 |
| States a course version passes through | 5, one of which is “in review” |
| Staff roles | 5, plus the student |
| Language models called by any code path | 0 |
| Courses of our own that we would sell your students | 0 |
Third
Your students, and your staff
Will my students actually use it?
We do not know. There are no customers, therefore no usage data, and anybody quoting you an engagement figure with no named school behind it has invented it. What we can tell you is what was taken out of the path: no app store, no download, nothing to install, and your name at the top rather than a supplier’s.
Do they need an account, or an app?
No app. It is a web page on the phone they already carry, and there is no store listing and no offline mode — that is a decision with a written record behind it, not a shortcut. They do sign in, and we are not going to describe that as no account: progress is kept on the server against a person rather than in the phone’s browser, where iOS caps script-written storage at seven days of browsing without your course being opened. It happens under your school’s own path, with your name and colours on it. A web address of your own is a short name we store and a verification step nobody has built.
What if I already have a website?
Keep it. This is not a website, it does not replace one, and it will not get you found by anybody who is not already your student. The way in is an access link an owner or an administrator generates and hands over — paste it into a message, into your own site, or print it for the notice board in the corridor. Nothing is emailed, because nothing in this product sends email. Nothing has to be migrated and nothing on your existing site has to change.
What can my staff break?
Five roles, and the useful question is what each one cannot do. An admin runs your people and your branding and cannot promote anybody to a role above their own. An editor writes lessons, uploads material and edits questions, and cannot publish a version or touch your people or the school itself. An instructor reads the course and sees how students are getting on and cannot edit anything. An analyst reads results and holds not one write permission. The school can never be left without an active owner — that is a database rule refusing the change, not a warning dialog — and suspending somebody ends their sessions at that moment rather than whenever their login happens to expire. Your staff and their roles has the full matrix.
Last
Ownership, and the awkward ones
Who owns what I send you?
You do. Your material stays yours and so does the course made from it; we take ownership of neither. We ask who holds the copyright before we convert a page, and if you do not hold it we say so and decline the work rather than take your money. Copyright and rights is the long version.
Can I take my course elsewhere?
Yes, and you should press on exactly how, because this is where the answer is worse than you would like. There is no export button in the software today — asking us to demonstrate one would catch us out, and you should hear that here rather than in the demonstration. What exists is a written undertaking: if you stop paying us, you get your converted course back in a form you can keep and take with you. If you leave says what that means in practice.
What happens if you go out of business?
It is the right question to ask a company younger than your newest instructor, and we would ask it too. There is no escrow arrangement and no fund behind us, and inventing one would be exactly the kind of claim this site exists not to make. What is true is that your material sits in your own space, we never take ownership of it, and the undertaking above is what you hold. Nothing is deployed anywhere yet — we are building this rather than running it — and that belongs on the page rather than in the third meeting. About says who we are and how the bills are paid.
If a question you have is not here, it is because nobody has asked it yet. Ask it, and if the answer is one we would rather not print, it goes on this page anyway.
Three of the answers above end at the same page, which is the one worth reading in full before you send anybody a chapter: your material
Send one chapter.
Not a demo of somebody else’s course. A chapter of yours, converted properly, so you can look at your own questions and your own page numbers and decide whether it is any good. Pick the chapter you know best. If it is no good, you have lost an email.
We will ask who owns the copyright before we convert anything. If you do not hold the rights to the material, we will say so rather than take the work.